Learning of Constraint Logic Programs by Combining Unfolding and Slicing Techniques

This paper discusses learning of Constraint Logic Programs using unfolding and slicing technique. The transformation rule for unfolding together with clause removal is a method for specialization of Logic Programs. Slicing is a program analysis technique originally developed for imperative languages. It facilitates the understanding of data flow and debugging. This paper formulates the semantics of a learning method of CLP programs, proves that the unfolding transformation preserves the operational and logical semantics, and combines the defined unfolding technique by applying slicing. A prototype learner of CLP programs which implements the above ideas is briefly described.